Further to the N. the III. Corps also had done most creditable,
if less spectacular work, during these two days of battle. The
$8th Div. on the right had cleared Marrieres Wood on the 3ist,
while the 47th Div. made progress towards Rancourt, repulsing a
hostile counter-attack. On the morrow these successes were
continued and completed by the capture of Rancourt and
Bouchavesnes, while the i8th Div. on the left of the corps in a
brilliant series of attacks seized in turn Priez farm, Fregicourt
and Saillisel, inflicting serious losses on the enemy, who afforded
a stubborn resistance.
Sept. 2 saw the completion of the successful operations of the
IV. Army in the battle of Peronne, the Australian Corps occupy-
ing Allaines and the III. Corps St. Pierre Vaast and Vaux woods.
The results of the battle were imposing enough even in mere
figures. In the period between Aug. 22 and Sept. 2 the IV.
Army's 9 divs. had engaged and defeated 23 hostile divs. and
taken from them over 23,000 prisoners, many guns and vast
quantities of material. The strong line of the Somme had been
turned and rendered untenable by sheer hard fighting in which
the attacking troops had shown themselves capable of meeting
and defeating the best of the German divisions which, thrown in
piecemeal and in the utmost haste and confusion as they arrived
MUlVi VtV.1'
pieceme
on the field, had been unable to hold for long even the strongest
natural and artificial defences.
General Results of the Battle of Bapaume- Peronne. In the
battle described above the British III. and IV. Armies, consisting
of five corps (23 divs.) in all, had fought and defeated the German
XVII. and II. Armies, consisting of five corps (46 divs.), had
forced them to fall back to a depth of from 6 to 13 m. on a front
of 28, and had captured from them a total of 34,250 prisoners
and 270 guns, without reckoning other material of war too various
to recapitulate. The whole area of the Somme battlefields, which
had cost the British five months' bitter fighting in 1916, had been
conquered in less than a fortnight; more than half the area
gained by the great German advance of the spring had been
recovered; the only good natural line of defence available for
the enemy to the W. of the Hindenburg system had been broken
asunder; and the moral and material superiority of the British
over the German fighting machine had become patent to the
world. (X.)
SONNINO, SIDNEY, BARON (1847- ), Italian statesman
(see 25^6). During the debates on Giolitti's Steamship Subsi-
dies bill in the spring of 1909 it was Baron Sonnino who con-
ducted the most vigorous attacks against the Government,
exposing the radical defects of the measure, and when Giolitti
resigned on Dec. 2 it was Sonnino who was called upon to form
a ministry, for the second time. But he did not enjoy the favour
of the still Giolittian Chamber, and his Cabinet was defeated
over the new shipping bill. On March 21 1910 he resigned,
again after roo days of office. He continued to take an active
part in the debates in the Chamber, and was a stern but just
critic of Giolittian political methods, although during the Libyan
war he generally abstained from opposition for patriotic motives.
In the autumn of 1914, after the death of the Marquis di San
Giuliano, the Premier Salandra assumed the Foreign Office for
a short time, but when he reconstituted his Cabinet on Nov.
5 he offered that portfolio to Sonnino, who accepted it. His
conduct of the Foreign Office was characterized by sincerity of
purpose, high principles, unswerving patriotism and a wide
knowledge of international poh'tics. He had not, moreover,
a free hand. He was still Foreign Minister, under Orlando's
premiership, during the Peace Conference, which he attended
as second Italian delegate from Jan. 18 to June ig 1919. On
the fall, however, of the Orlando Cabinet (June 19 1919)
Sonnino retired into private life. The irritation of the whole of
Italy against the policy of the Allies towards Italy at the
Peace Conference reacted to some extent against the nation's
representatives at Paris, and Sonnino himself came in for a
large share of unpopularity, although the more intelligent and
better informed part of public opinion realized the great diffi- >
culty of his task and the insufficient support afforded him by
Orlando, as well as the value of his actual achievements. He
did not stand for Parliament at the elections in Nov. 1919,
but was subsequently made a senator. In spite of what was
regarded as his failure to overcome the obstacles of the Peace
Conference, he enjoyed the reputation of being the greatest
Minister for Foreign Affairs that Italy had had since Cavour,
with the possible exception of Crispi, while as a financier he :
ranked very high. He was also a man of wide reading and cul-
ture, and a distinguished Dante scholar and bibliophile.
SOROLLA Y BASTIDA, JOAQUIN (1863- ), Spanish
painter (see 25.434), was engaged for practically the whole of
the decade 1910-20 on work for the Hispanic Society of America.
It includes a series of portraits of Spanish writers, and a " Pano-
rama of the Forty-nine Provinces of Spain " consisting of forty-
nine immense compositions, each representing views, costumes
and customs of a different province. This great undertaking
was completed before paralysis brought the artist's painting to
an end. Important exhibitions of his work were held at the
Graf ton Galleries, London, 1908; in New York, 1909; in Chicago,
1913; and he was represented by two typical works in the 1920-1
Exhibition of Spanish Paintings at Burlington House.
See Hispanic Society of America, Eight Essays on Joaquin Sofolla
y Bastida, 1909; A. de Beruete y Moret, Sorotta y Bastida, 1920.
526
SOUND
SOUND (see 25.437). The increase in our knowledge of the
subject of acoustics (the science of Sound) during recent years
has been largely associated with the war conditions which pre-
vailed from 1914 to 1918. As a consequence of the war the
development of this science has been abnormal, 1 and research
has been directed towards the rapid realization of practical
acoustic devices and methods for immediate use in warfare,
both on land and sea. A general survey of the work done shows
that the advances consist of applications of well-established
principles, rather than the discovery of new phenomena. Gener-
ally, the observations made have proved to be in accordance
with previous theoretical investigations, mainly due to the late
Lord Rayleigh. 2 This war work falls naturally under two
headings, viz. (i) the detection and perception of direction
of sounds in air, and (2) the detection and perception of
direction of sounds in water. Theoretically, these two prob-
lems have much in common, but, practically, there are im-
portant differences which make it desirable to treat them in
separate sections. A special section (3) is devoted to the
important advances recently made in auditorium acoustics, and
the remaining section (4) deals briefly with miscellaneous out-
standing features of modern work on sound, not essentially mili-
tary in character.
i. DETECTION AND PERCEPTION OF DIRECTION OF
SOUNDS IN AIR*
Detection. The human ear itself is a remarkably sensitive
detector of the air vibrations which constitute sound. It is
still much superior in this respect to any mechanical device
which has yet been produced for recording the vibrations
visually. Thus the perception of feeble sounds of necessity
depends upon the limitations of audibility, either indirect
listening, or with the ear aided by the intervention of an electri-
cal device such as a microphone. The audibility of a feeble
sound can be very largely augmented by making use of the
principle of resonance, provided that the sound itself approxi-
mates to a pure tone. This can be secured, for example, by
the use of a Helmholtz resonator applied to the ear in the case
of direct listening, and in addition, by tuning the diaphragm
receiver when microphonic listening is adopted. It has hap-
pened fortuitously that one of the chief sounds in air which it
is important to be able to detect, viz. those emitted by air-
craft, do contain predominant notes which enable the applica-
tion of resonance, as above indicated, to increase largely the
range of audibility. Typical predominant frequencies (appar-
ently due to engine exhaust) are given in the following table,
which relates to the engine running at the usual speed:
Aeroplane engine
Frequency (vibrations per
second)
S. E. 5
130
R. E. 8
90
F. E. 26
70
Avro
90
Gotha
80
1 Owing to the abnormality of the conditions, it is impossible to
follow the usual practices in writing the present article. Experi-
ments on sound, with military ends in view, have been carried out
in nearly all the belligerent countries. Comparatively few of the
results have found their way into the recognized scientific journals,
largely by reason of the secrecy which is still frequently enforced by
the various Governments, under whose control most of the work was
done. In the circumstances it is not safe to attempt to assign credit
to particular investigators, nor is it possible to give adequate refer-
ences. The present article has been drawn up, therefore, upon
broad general lines which, since they fulfil censorship conditions,
form necessarily a by no means complete survey; and names have
been avoided as far as possible.
" Lord Rayleigh's work is contained in his Collected Papers (No.
6, 1920). His contributions were numerous between 1911 and 1919,
when he died.
1 The information contained in this section is largely drawn
from a manual entitled Development of Sounds, kindly placed at
the writer's disposal by the British Munitions Inventions Dept.
The following frequencies have been detected in the sounds from
the Maybach engines of a Zeppelin airship:
Slow speed 27, 54, 108, 135, 243.
High speed 57, 114, 171, 228.
The operation of the Doppler e/ect, arising from the relative
motion between the aircraft and the observer, prevents the
possibility of the identification of the machine by means of the
observed frequency, this being liable to change by as much as
20%, according to the speed and direction of flight. An inter-
esting observation which has been constantly made is that the
notes of low pitch continue to be heard at ranges where those of
high pitch have ceased to be audible. This is in accordance
with the theoretical expectation that damping increases with
frequency.
The determination of the direction whence a sound arrives is
theoretically possible by a variety of methods dealt with below,
several of which have been tried in aircraft localization.
(a) Binaural Listening,* Lord Rayleigh's experiments (Collected
Papers, vol. 5, p. 347) have shown that low-pitched sounds are
determined in direction by the observation of the phase difference
between the vibrations arriving at the two ears. This principle has
been applied in direction-finding, and the effect has been exaggerated
by increasing the distance between the two points of reception.
The sound is received by two equal trumpets or horns rigidly con-
nected together and capable of rotation about an axis perpendicu-
lar to the line joining them. Separate and exactly equal tubes lead
from the trumpets to the two ears, respectively, and the apparatus
is rotated until the sound under observation appears to come from
directly in front. The line joining the sound receivers is then per-
pendicular to the incident sound stream. An alternative method
which dispenses with the necessity of rotating the apparatus is
to use a compensator or phase-measurer, which consists of tubes,
adjustable in length, inserted between the sound receivers and the
appropriate ears, so as to provide a path difference equal to that
between the distant source of sound and the two receivers. Adjust-
ment of the tube lengths is made until the impression received is
that the sound is neither to the right nor to the left, and the deter-
mination of direction is then a matter of simple geometry. In prac-
tice the compensator is graduated to give direct angular readings.
The practice of binaural listening has verified theoretical conclu-
sions in several important respects. It has been found that it is
easier to perceive the direction of a mixed sound, or noise, than a
pure note. Apparently it is necessary that the wave train should
contain more or less isolated special characteristics whereby the
phase difference can be readily appreciated. In the regular sine
wave corresponding to a pure tone each vibration is exactly like
those which immediately precede and follow it, and the ears are
unable to identify corresponding displacements. It is apparently
also necessary for successful binaural listening that the two por-
tions of the incident wave which enter the two receivers should be
free from subsequent distortion ; in particular, that the sound
receivers should be as nearly as possible non-resonant for the vibra-
tions in question. Any amplification of the sound which depends
upon resonance, therefore, such as the use of Helmholtz resonators
already referred to, is incompatible with efficient direction-finding
by observations of phase difference.
The construction and arrangement of the receivers used has varied
very much in practice. As a typical system, that commonly used in
the British army may be quoted, namely, circular cones, 2 to 4 ft.
long and of semi-angle 20, as receivers, placed about 7 ft. apart
a distance which proved to be sufficient for attaining nearly the
maximum practical accuracy of setting.
The method is subject to many errors, chiefly those arising from
the motion of the sound source, refraction due to temperature ine-
qualities in the air, and the effect of winds. The necessary correc-
tions are tabulated for use in practice.
(b) Sound Mirrors. Some success has been attained in direction-
finding by means of concave sound reflectors. The chief limita-
tions have arisen from the question of size, and, consequently, of
4 This method of perception of direction has been largely used
also in a connexion which scarcely justifies treatment in a separate
section. The ffophone is an instrument for direction-finding of
sounds proceeding through the earth, and its particular use during
the war was for localizing the sounds of picks, etc. used in tunnelling
and land mining. It consists of two hollow boxes connected by equal
tubes to a stethoscope arranged so that the sounds proceed from the
two boxes to separate ears. The boxes are laid upon the ground a
few feet apart, and moved about until the sounds of the pick appear
to come from straight ahead. It is then known that the sound
source is on a line perpendicular to that joining the two geophone
receivers, since the sounds arrive through the earth in synchronism.
By combining several pairs of geophones separated by considerable
distances, the actual position of the pick can be estimated, for it
lies at the intersection of the several perpendiculars above specified.
SOUND
527
portability. In optics the size of mirrors commonly in use is very
great in comparison with the wave-lengths of the light ; in the cor-
responding problem in acoustics it is almost impossible to make
them so; and yet this is a necessary condition for the geometrical
laws of reflection to apply with accuracy. In the largest sound mir-
rors perhaps 20 ft. in diameter the size is at most only a few
wave-lengths for the aircraft sounds under investigation, with the
result that the image of a distant sound obtained at the focus proves
to be an area much larger than that corresponding to optical calcu-
lations. There is therefore no advantage secured by making the mir-
ror parabolqidal instead of spherical, and considerable roughness of
the surface is not detrimental. The mirrors were usually made of
concrete, and listening was effected either by means of a small horn
receiver placed in the focal plane and connected by a tube to the
ears, or by means of a microphone placed in a similar position. If, as
was more usual, the mirror was fixed, the direction of the sound
source could be found by determining the position of maximum inten-
sity in the focal plane. It may be noted that in this method of direction-
finding amplification is obtained on account of the area of the
mirror, and that further augmentation is attainable by using resona-
tors, to which the same objections do not apply as in binaural listen-
ing. The accuracy of the determinations vary very much with fre-
quency, being much greater for notes of high pitch than for low, as
would be anticipated from considerations of wave-length.
(c) Interference and Diffraction Methods. There have been many
attempts to apply the principle of interference as a substitute for
binaural listening, i.e., by ultimately mixing the sounds entering the
two receivers, instead of leading them to different ears, and adjust-
ing the compensator until the total sound heard is as loud as possible.
Theoretically this will occur when there has been provided in the
compensator a difference of path equal to the path difference out-
side the receivers. The method has not proved very successful, for a
variety of reasons, some of which are obscure. We shall not elab-
orate them here.
On the other hand, remarkable results have been obtained by the
application to sound waves of a phenomenon well known in the
diffraction of light. A small distant source of light gives in the mid-
dle of the shadow of a small circular obstacle a luminous region,
called the " white spot," arising from the diffraction of light round
the edges of the obstacle. The same phenomenon is observable in
sound under suitable conditions. Thus a large horizontal disc, at
least 20 ft. in diameter, and made of material which either reflects
or absorbs sound, will give below itself a sound shadow of a sound
source, such as an aeroplane, above it. Near the centre of the shadow,
in a position depending on that of the source, there is a region where
the sound heard is comparatively loud in many cases much louder
than it would be if the disc were absent. The relation between the
direction of incidence of the sound and the position of maximum
intensity has been calculated, and the method provides, perhaps,
the most reliable means of perceiving the direction of air-borne
sounds.
(d) Sound Ranging. This special military aspect of the localiza-
tion of sound sources, viz. those arising from gun-fire and shell
bursts, is dealt with in the article RANGE-FINDERS.
2. DETECTION AND PERCEPTION or DIRECTION OF
SOUNDS IN WATER'
Of all the methods practised for the detection of submarines
that depending on the sounds which they emit has been of the
widest application. The question of. detection has been, of
course, of nearly equal importance in the opposite sense, viz.
the hearing of surface ships by the crew of a submerged subma-
rine. The sounds created in the sea by a screw-propelled ship
are of a very complicated character, arising partly from the
interaction between the propeller and the water, and partly from
the vibrations of the machinery which are transmitted through
the walls of the ship into the sea. They vary greatly from ship
to ship, even of the same class; and, in the later stages of the war,
submarines had been constructed which, when cruising sub-
merged at certain slow speeds, emitted practically no noise at all.
In many ways the detection of submarines in the sea is more
difficult than that of aircraft in air. Normally, listening in air
takes place at stations which are fixed; in submarine listening
he stations were most frequently ships, which for tactical reas-
ns connected with their safety, had to be constantly on the
nove. Their own machinery noise and the acoustic disturb-
nces arising from their motion through the water were very
l The following publications should be consulted, although, for
"asons already given, they form by no means adequate references :
. C. Hayes, Engineer (1920), p. 491 ; C. V. Drysdale (Kelvin Lee-
re), Journ, I.E.E. (1920); W. H. Bragg, Submarine Acous-
cs," Nature, July 1919; F. L. Hopwood,
Mature, Aug. 1919.
1 Submarine Acoustics,"
apt to drown the noises proceeding from more distant sources.
The noise of the sea, too, even in weather not at all stormy,
interfered greatly, and the range at which a submarine could
be heard varied much from day to day. A serious additional
limitation was that recourse could not normally be had to the
reflection of ordinary sounds (as is possible in air) chiefly by
reason of the great size of the necessary reflectors. For the
speed of sound in sea water is more than four times that in air,
so that the wave-lengths are larger in the same ratio. This
necessitates a corresponding increase in the linear dimensions of
the sound mirror, if equal efficiency is to be obtained.
Hydrophones. Hydrophones, or under-water sound detectors,
were already in use before the war for signalling purposes, being car-
ried by ships for listening to submarine bells operated by Trinity
House as warnings in foggy weather. They consisted of small, metal,
water-tight cases of which one face was a metallic diaphragm oper-
ating an enclosed microphone. The electrical disturbances of the
microphone caused by any vibration of the diaphragm arising from
sound pressure waves in the sea, were conveyed to telephone receiv-
ers on the ship, where listening took place. It was usual to suspend
the hydrophones in water-filled tanks attached inboard to the outer
shell of the ship, which, owing to the fact that steel in water trans-
mits sound almost completely, does not diminish appreciably the
intensity. Normally the hydrophone diaphragm was tuned so that
its natural frequency in water 2 approximated to that of the sig-
nalling bell, and so that increased range could be secured by depend-
ing on resonance.
The earlier hydrophones used for naval purposes were of much
the same type, although the resonant diaphragm proved to be by
no means an unmixed advantage. All sounds containing a com-
ponent corresponding to the diaphragm frequency were distorted in
reproduction, and what was gained in sensitivity was liable to be
lost in the difficulty of recognition, or, in other words, failure in
discrimination between genuine noises due to a submarine and other
noises inevitably present in the sea. Appeal to resonance is only
really advantageous when the sound under observation has a pre-
dominant note, as in the case of an aeroplane; and submarines do
not display this characteristic. Ultimately hydrophones of a non-
resonant character came to be preferred, and were frequently used
in practice. These consisted most usually of enclosures made of
rubber, sufficiently thick to withstand the pressure of the sea at
the usual depth (about 15 ft.), and having natural frequencies below
the limit of audition.
An alternative type of hydrophone consisted merely of a hollow
enclosure without a microphone, sometimes with a metallic dia-
phragm, and sometimes simply a rubber tube, filled with air and
connected by long tubes to stethoscopes applied to the ears. These
are operated by the transference of the pressure vibrations from the
sea to the air cavity and thence to the ears. Electrical hydrophones
have the advantage over non-electrical ones that their sensitivity
can be readily augmented by various means, e.g., bv the use of
thermionic amplifiers (see WIRELESS TELEGRAPHY).
In cases where the hydrophones had to be used by ships in motion,
they were sometimes fitted into the hull of the ship; or themselves
consisted of fish-shaped bodies towed at a considerable distance
behind the ship. The former precaution, i.e. making the shape
stream-like, aimed at diminishing the vibrations created by the
passage of the hydrophone through the water; the latter had in
view the partial elimination of the disturbances arising from noises
in the towing ship. Even so, it frequently became necessary to stop
the engines temporarily, and listen with the towed hydrophones
while the momentum of the ship continued to carry it forward.
This proved to be only feasible at comparatively slow speeds.
Directional Hydrophones. All the hydrophones so far described
are of a non-directional character, i.e. the intensity of the sound
heard in them is practically independent of the orientation of the
sensitive receiving diaphragm with respect to the position of the
source of sound. The limitations of dimensions necessitated by coa-
siderations of portability, etc., are such as to render the instruments